Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!


Forgot your password?

BitTorrent Closes Source Code 390

An anonymous reader writes ""There are two issues people need to come to grips with," BitTorrent CEO Ashwin Narvin told Slyck.com. "Developers who produce open source products will often have their product repackaged and redistributed by businesses with malicious intent. They repackage the software with spyware or charge for the product. We often receive phone calls from people who complain they have paid for the BitTorrent client." As for the protocol itself, that too is closed, but is available by obtaining an SDK license."
This discussion has been archived. No new comments can be posted.

BitTorrent Closes Source Code

Comments Filter:
  • Who cares? (Score:5, Interesting)

    by Rix ( 54095 ) on Wednesday August 08, 2007 @10:50PM (#20165275)
    The company that owns the BitTorrent trademark is not the arbiter of the protocol or anything else. Do they even own that trademark?

    Note that they opposed the addition of encryption, and they were completely ignored. BitTorrent, the company, is entirely irrelevant.
  • Bittornado (Score:3, Interesting)

    by urikkiru ( 801560 ) on Wednesday August 08, 2007 @10:55PM (#20165325) Journal
    http://www.bittornado.com/ [bittornado.com]

    There, that should tide us over for a while.
  • Re:So.... (Score:5, Interesting)

    by pilot1 ( 610480 ) on Wednesday August 08, 2007 @11:12PM (#20165433)

    What's the name going to be for the upcoming auto-encrypted open-sourced fork of Bittorrent?
    This is where it could get ugly. uTorrent is the most popular client, at least according to the article, and it's closed source. If the protocol is forked and modified enough to be incompatible with the older protocol versions, there's going to be some fragmentation. Anyone using uTorrent wouldn't be able to connect to people using the new protocol. uTorrent users would have to switch to a new client if its developers refused to update its protocol. Or worse, uTorrent users might continue to use uTorrent while everyone else uses the new protocol, causing nasty fragmentation.
  • by jmorris42 ( 1458 ) * <jmorris@[ ]u.org ['bea' in gap]> on Wednesday August 08, 2007 @11:12PM (#20165437)
    > Yes - they're maintaining the Open Source client... with a protocol that they hint they will be leaving behind.

    One difference. They don't operate any of the servers people actually use. Unless they can convice the server operators (most of whom they can't legally even admit exists, which will make negotiations somewhat awkward) to adopt their closed protocol, who will notice any optional dead protocols their 'official' but little used client supports?

    At this point someone simply needs to write up a formal documentation of the protocol as it currently exists and submit it to the W3C, at which point the wire protocol is pretty much settled. And go ahead and pick a new anme because you can bet your last dollar they will pull the trademark crap the second they realize they are being written out of the picture.
  • by PaintyThePirate ( 682047 ) on Wednesday August 08, 2007 @11:14PM (#20165449) Homepage
    It's still a matter of what you, as a developer, can do with the protocol. Obviously, part of the SDK licensing agreement will be that you can't just publish it for the world to see, or be allowed to incorporate it into an open source project (or probably even a third party closed source project).
  • by _Sprocket_ ( 42527 ) on Wednesday August 08, 2007 @11:35PM (#20165581)

    At this point someone simply needs to write up a formal documentation of the protocol as it currently exists and submit it to the W3C, at which point the wire protocol is pretty much settled. And go ahead and pick a new anme because you can bet your last dollar they will pull the trademark crap the second they realize they are being written out of the picture.
    There seems to be echoes of SSH in this story. Granted - the history of SSH involves some distinct differences (for example, Tatu Ylönen submitted SSH to the IETF as a standard which set the grounds for "SSH" becoming hard to restrict despite SSH,Inc.'s annoyance at the "OpenSSH" name). But one can't help to wonder if this will pan out the same way; the last BitTorrent OSS release becoming a springboard for continued development that competes if not completely overshadows the originator's own efforts.
  • I can only hope... (Score:4, Interesting)

    by BlueCoder ( 223005 ) on Wednesday August 08, 2007 @11:55PM (#20165721)
    That my fellow community developers will take this opportunity to drop the BitTorrent protocol. Time to develop something better.

    It's time we address it's critical failure... that you can see which IP's are trafficking in which files. There has to be an obscure way in which people can just exchange data blobs. Where the blobs are interleaved or multiplexed with data of other files and you don't know and can't know with all practicality what a particular blob contains until you finally collect enough blobs to reconstruct your data file. There are more blobs to be collected for a particular file for data redundancy but you only need to collect so many of them to recreate the data set. Meanwhile sure you downloaded more data then you needed to for that particular file but all the blobs you downloaded are still in demand from other people because of their relevance to other data sets. And you can safely continute to server those files because you don't necessarily know what multiplexed data they contain. Blobs also mutate and remix over time as to which combined data they contain.
  • by spoonboy42 ( 146048 ) on Wednesday August 08, 2007 @11:58PM (#20165759)

    From the article itself, it appears that, since acquiring uTorrent, a closed-source C++ BitTorrent client for Windows, Bittorrent, inc. has decided to keep it closed source, and also to make it the new "mainline" BitTorrent. The old "mainline" client, which is open-source, written in Python (with wx for the graphics) and is generally cross-platform, last I checked, will continue to be maintained as a "reference implementation", but might not always track the latest protocol updates to uTorrent. Full documentation on the protocol will apparently come with an "SDK license", which they claim is "easy to get".

    Well, first of all they ARE doing a few things that contradict the spirit of free software. Their main client app will be closed source, and although the reference implementation will apparently continue to be free, protocol docs require you to acquire a special license. A few years ago, these moves would have tightened Bittorrent inc's grip on the world of bt clients in general.

    Now, however, the landscape is different. I can't produce statistics for all torrent users in general, but when I take a look at my peers in my preferred client, KTorrent [ktorrent.org], there seems to be a near dead-heat for most popular client between uTorrent and Azureus [sourceforge.net] (also open source), with certain alternative clients like Transmission [m0k.org], Bitrocket [bitrocket.org], and KTorrent [ktorrent.org] making frequent appearances, as well (and all 3 of those examples? also open source). Although uTorrent certainly remains a big player, it doesn't confer upon BitTorrent, inc. the ability to dictate major compatibility-breaking protocol changes by fiat. The fact that the main implementation of BT was open source to start basically stops things from being ruined by more restrictive licensing now.

  • by Anonymous Coward on Thursday August 09, 2007 @12:07AM (#20165821)
    AFAIK, if a protocol is not patented, the only protection available
    is trade secret, which evaporates when published. I'm not sure they
    can pull it back now.

    It's similar to the reason you can't really protect the definition of a
    programming language, even though lots of companies act like you can
    and the industry largely plays along.
  • by Max Littlemore ( 1001285 ) on Thursday August 09, 2007 @12:11AM (#20165857)

    While we're at it, let's point out how wonderful some of those tags are.

    This story is tagged "lame" and "bastards" among other things. So yeah, if I'm interested in looking up info on OSS software being closed, I'll be sure to look for articles tagged "lame". That imediately makes so much sense to me, and you guys clearly know what good tagging's all about. Tagging's a great way of expressing opinions on entire stories without having to own up to them. You don't even have to have to LEAVE A FUCKING COMMENT WITH A USER NAME.

    C'mon, at least post AC, dickheads.

  • by Kadin2048 ( 468275 ) * <slashdot.kadin@NoSPaM.xoxy.net> on Thursday August 09, 2007 @12:58AM (#20166087) Homepage Journal

    The problem being that when one company has near monopoly, and in the eye of the public is indistinguishable from the product, they can close source, then change the specs (even if the spec is published), and the open source alternatives won't be able to compete.
    They have a monopoly ... how, exactly?

    People use Bittorrent -- or more specifically, many people use uTorrent -- to connect to public BT trackers and to other people running similar client programs. Bittorrent (the company) doesn't control either. In fact, I don't think that Bittorrent-the-company's "reference implementation" is particularly popular for trackers, and they're really where the marketshare matters.

    I don't think that the majority of bittorent (the protocol) users are just going to bend over and throw away the software that they've liked, just because Bittorrent (the company) decides it would be cool to produce a new, ad-laden, DRM-using, Hollywood-mogul-approved version of their software, that breaks compatibility with older versions. In fact, I strongly suspect that the trackers which drive the more popular torrent aggregation sites would refuse to recognize such a "broken" implementation, and would instead favor free implementations (old versions of uTorrent, Azureus, etc.).

    What's happening here is that Bittorrent (the company) has become fully decoupled from bittorrent (the protocol). They have very little leverage over the latter; about all they have is the rights to the name "Bittorrent," and the 'reference implementation,' which won't be worth its weight in electrons once they start messing with it.

    The comparisons to Microsoft and RTF aren't really apt, because Microsoft had a way they could easily control the format -- they just made future versions of Word produce output that was incompatible with other vendors' software. But Bittorrent can't really do that, because a bittorrent client is only useful insofar as it can communicate with the swarm. As long as the trackers that drive the most popular torrents (which, let's face it, are the illegal ones; warez and movies) don't start using the new/broken protocols, it seems unlikely that a broken protocol would gain traction.

  • bla bla bla blobs bla bla bla

    bla bla finally collect enough blobs bla bla

    downloaded more data then you need bla bla bla
    Worst. Protocol. Ever.
    And that's only skimming your description.

    Besides, not being able to preview files will pretty much make it useless for anything mainstream. Like pirating crap. So, if this protocol is never used for piracy, it will never need such insane protection from the MAFIAA because it will never blip on their radar. Oh, it can be used for other things, like downloading Linux ISOs? BT already does that. Secure file transfer? LOL Traffic analysis foiling? Tor. What else?

    You are coming to a sad realization: Cancel or Allow?
  • Re:So.... (Score:4, Interesting)

    by Antique Geekmeister ( 740220 ) on Thursday August 09, 2007 @02:33AM (#20166497)
    Yeah, they're trying to act like Dan Bernstein (with daemontools and qmail) and prevent forks that do things in ways they don't like. Some of that desire may be legitimate, but one real desire is probably the desire to avoid encrypted transfers becoming common place and leavingn them vulnerable to governmental complaint about un-tappable data transfers.

    I can easily picture the various motion picture and software copyright lawyers sending a few dark glasses wearing "lawyers" to explain "nice little business you got here, I'd hate to see anything happen to it" to encourage Bittorrent both to avoid providing encrypted transfers and to add "load monitoring" features that ease tracking. I'm not saying this is sure to happen, but with the source closed, it wouldn't take much to add hooks to report specific downloads to the mothership.
  • by kestasjk ( 933987 ) on Thursday August 09, 2007 @04:46AM (#20167041) Homepage
    Seems like this'll only split the bittorrent protocol, there's a fairly wide variety of clients out there and the only thing that held them together was the official protocol. Azureus has been making small breaks even with the official protocol around, so now it'll probably split. The question is which client will the other ones follow, now that BitTorrent have given up their niche in true XFree86 style.
  • by Captain Splendid ( 673276 ) <capsplendid AT gmail DOT com> on Thursday August 09, 2007 @04:47AM (#20167047) Homepage Journal
    I gotta say, as clunky as Azureus has been, they've obviously been working hard on the thing, because it works so much more smoothly now. Even with a few torrents running, I don't get huge CPU grabs like I used to, and the overall feel of speed is definitely improved.
  • by Toby_Tyke ( 797359 ) on Thursday August 09, 2007 @06:20AM (#20167491) Journal
    FUD is pretending that bit torrent is NOT used mostly for piracy. Take a look on piratebay or mininova. The vast majority of torrents on there are for copyrighted material that the uploaders have no legal right to share. All the people downloading Linux distros probably don't even come close to the numbers downloading movies. Admittedly, I'm not aware of any detailed research on the issue, but the evidence available clearly indicates that the most popular use of bit torrent is illeagle distribution of copyrighted works.

    That's certainly all I use it for.
  • by Stiletto ( 12066 ) on Thursday August 09, 2007 @08:56AM (#20168377)
    So, release yours as open source, assuming it works. There's clearly demand for an open protocol such as this, and one that "did things right" and outperformed BitTorrent would probably become quite popular.
  • Monetizing the SDK (Score:3, Interesting)

    by Valdrax ( 32670 ) on Thursday August 09, 2007 @09:24AM (#20168711)
    Bit torrent have made a closed source client their mainline client, and have decided to fortify their rights to the protocol too (its closed, but an SDK can be requested).

    Correction -- their SDK can be *paid for*.

    I beginning to think that the whole point of acquiring the most popular closed source client was to allow them to close and charge for the SDK. The counterpoint to this argument is that if any one open source P2P grits it teeth and pays whatever fee they're going to charge open source clients, then their implementation becomes the new reference.

    A lot of people here are talking about how the Mainline client has once again aggressively pursued irrelevance, but uTorrent's marketshare is going to be nigh impossible to unseat unless they do something self-destructive like removing a popular feature they don't like (like encrpytion). They have a really good chance of dictating the development of the future of the protocol with that client in hand.

    I think that this finally explains the reason for the buy-out and the lack of open source. I had previously thought it was due to uTorrent's original developer's dislike of open source, but it may have more to do with control and with monetizing the SDK.
  • by Crayon Kid ( 700279 ) on Thursday August 09, 2007 @09:42AM (#20168937)
    If you want to see where this may be headed, take a look at the DirectConnect situation, or many other popular P2P protocols.

    DC is also a popular P2P protocol and it started as a closed application whose protocol was reverse engineered. Later attempts to retake control were futile and nowadays there's no such thing as an "official" DC protocol, only several different client software making it on sheer popularity. Just like BT, some of them add new features and sometimes they're borrowed by the others and so on.

    Think of IRC too. It also doesn't have an "official" specification, there are all these servers and clients and so on. At least there were some RFC's at some point, which is more than can be said of other P2P protocols.

    So it seems to be a "normal" situation with P2P to not have a standard protocol and for it to evolve on server/client software popularity alone.
  • by Doctor O ( 549663 ) on Thursday August 09, 2007 @09:54AM (#20169095) Homepage Journal

    No possibility of adding multicasting?? Wtf does this even mean?

    Sorry to ruin it for you, but if you don't even know what that means, you're not in a position of judging *any* protocol.

    Another hint for you: Just because it seems to work well for you on a small scale doesn't mean it is a good protocol (i.e. scaling well, little overhead, easily extendable, etc.).
  • Re:rtorrent pwnz (Score:3, Interesting)

    by Deagol ( 323173 ) on Thursday August 09, 2007 @01:54PM (#20172437) Homepage
    I've benchmarked a handful of select programs with various optimizations (FreeBSD, not Gentoo). There's certainly no general/golden rule. I do agree that often (maybe 50% of the time) that -Os is a very good option for some small libraries and programs. However, there's a whole world out there beyond -O optimizations (which I know are just short-hand pre-sets for collections of various flags).

    What someone needs to do is set up an automated build system that uses "-fprofile-generate/-fprofile-use" -- especially for base system, multimedia, and compression libraries. I got something like a 12% speedup in using that for mplayer (or was that mpg123...) in benchmark mode, as opposed to stock ports compiler options under FreeBSD/amd64 (which, I believe, is "-O2 -fno-strict-aliasing", though many individual ports tweak those to the extreme). Granted, each program would need a decent corpus of representitive "work" to do for this method to be effective, but it could be done on port-by-port basis.

    If someone wants to rock the world of Gentoo (and has more free time than sense), they should fork the project and incorporate the "profile" method I mentioned above *and* ACOVEA [coyotegulch.com] into an automated build system. Now *that* would be impressive. :)

To avoid criticism, do nothing, say nothing, be nothing. -- Elbert Hubbard